Can AI Review Improve Paper Drafting? An Empirical Study on 20 Computer Architecture Submissions
Di Wu · 2026
Read Can AI Review Improve Paper Drafting? An Empirical Study on 20 Computer Architecture Submissions by Di Wu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-01013.txt.
Description
Wu asks if AI review can help authors draft better papers in this 2026 study. AI written papers are flooding in, and peer review strains under the load. Venues have begun to use AI review, quietly in places. Worries about secrecy, quality, and fairness run high, with no consensus in sight. Wu sets that fight aside and tests a practical use.
His case study covers 20 computer architecture submissions with varied lineage. He built a tool that drafts structured AI reviews of a paper. It picks reviewers from a diverse pool, then clusters and ranks their notes by shared ground and weight. Human comments can be aligned with AI ones for scored comparison. The study finds AI review covers a large share of the issues humans raise. It also flags issues the humans missed. The claim is narrower: as a drafting aid, its reviews can catch real flaws early.
Catalog record
| CREATOR | Di Wu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 55 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.01013